2 Nuclear Stocks Chasing a $7 Trillion Opportunity

McKinsey estimates AI-driven global data center spending could reach $7 trillion by 2030, framing a large potential demand tailwind for nuclear power suppliers. The article highlights Oklo and NuScale Power as two SMR stocks positioned to benefit from the need for reliable, clean baseload power for AI data centers. The piece is largely promotional and comparative, with no new company-specific financials, but it reinforces a constructive investment narrative for the sector.

Analysis

The first-order read is that both names are a levered bet on the same bottleneck, but the second-order winner is the one with the cleaner route to financed deployment, not the best reactor concept. In this phase of the cycle, capital markets will reward “credible near-term milestones” far more than ultimate unit economics, so price action can stay disconnected from commercial reality for quarters. That favors momentum in the higher-beta name, but it also means any delay in licensing, site selection, or utility partnerships can re-rate these stocks sharply lower.

The more important competitive dynamic is that data-center power demand is likely to be solved in layers: interim gas, behind-the-meter generation, then longer-duration nuclear. That means SMR equities can gap on narrative yet still lose share of wallet to simpler bridge solutions if hyperscalers prioritize speed-to-power over decarbonization. A surprise here is that utilities and gas-turbine suppliers may be the near-term beneficiaries, while pure-play SMR companies become the “optionality trade” rather than the core infrastructure trade.

The market is probably underestimating the financing and execution drag. Even with strong demand signals, these projects are capital intensive, regulatory-heavy, and likely to see repeated schedule slippage; that creates a wide dispersion of outcomes and makes valuation multiples unreliable in the near term. The tradeable catalyst set is not earnings but contract announcements, NRC-related approvals, and evidence that hyperscalers are willing to sign take-or-pay power agreements.

Consensus may be overpaying for the AI tie-in and underpricing dilution risk. The longer this stays pre-revenue or pre-cash-flow, the more likely follow-on offerings become part of the equity story, especially after headline-driven rallies. The cleanest expression is to own the company with better optionality to specific customer wins, while fading the one whose equity is more exposed to broad thematic enthusiasm and slower utility-scale conversion.

Key Decisions for Investors

• Long OKLO / short SMR for the next 3-6 months as a relative-value expression on faster commercialization and higher narrative elasticity; size modestly because both names are high-volatility and can overshoot on contract headlines.
• If long SMR, use call spreads rather than stock ahead of licensing or partnership milestones; the convexity is attractive, but dilution and schedule risk make outright equity ownership poor risk/reward after sharp rallies.
• Avoid chasing both names at once; wait for either a 15-25% pullback or a confirmed new contract/NRC catalyst before adding, since the trade is more event-driven than fundamentals-driven in the next 1-2 quarters.
• Pair long OKLO with short a utility-basket or gas-turbine beneficiary if you want to isolate ‘direct AI power’ optionality from the broader power-stack trade; this reduces macro power-demand exposure and focuses on nuclear adoption alpha.
• Take profits into contract-announcement spikes: trim 30-50% of any position after a 20%+ gap higher, because these stocks are likely to mean-revert once the market prices in execution risk and financing needs.